-first of all: something concerning the first Run6 bhabha calibration (because there wasn't a ops meeting this monday), later some news from the rolling bhabha calibration
-web page of the First calibration: first run 6 bhabha calibration

rolling bhabha calibration issues

what happened

-accumulated 3e6 bhabhas by the end of last week
-calibration startet, constants have been calculated, cut out works, output of non validated channels works, everything went fine...until storing the constants
-storing crate 3 failed: "Failed to store crate 3: CalErr: failed, code 42, too many channels failed validation"
-which means: crates 0-2 got stored, but 3-9 not, temporal db content was not deleted --> the crates got stored dozens of times up to now
-most probably the reason for that mess: Wrong constants file loaded!
-loaded constants file is the same like the one, loaded in the regular condition last July, except:
original file:
# Start/end of validity
18Jul2006 23:59:59 utc
13 109 43702007 5742 6622 1.00942
13 110 4370200b 5743 6287 0.981305
13 111 43802003 5541 4741 1.02933

loaded file:
# Start/end of validity
14Jan2007 23:59:59 utc
13 109 43702007 5742 6622 1.00942
v13 110 4370200b 5743 6287 0.981305
13 111 43802003 5541 4741 1.02933

-probably during editing the validity date, this "v" got into this line
-this "v" somehow screwed up the loading without giving an error message to the logfile, Cdb constant for 13:110 is 0
---> I only checked the logfile for errors and some random samples to see if the constants are the same, which obviously was not enough :-(

what can we do?

1) switched off the code --> worst solution!
2) modify the CalType object to let a defined number of non validated channels pass --> not really a solution!
3) load the new bhabha calibration to both conditions, delete temporal db content now and after the loading --> good solution, since it will take at least another week until it's loaded
4) load the corrected file into EmcBhabhaTypeTest condition asap and see if it finishes the calibration --> good solution as well, may result in problems with the first 3 crates in the next calibration...

-comparison between online and offline calibration is not possible now (in any case)!
-could compare the successful stored crates of the first storing and the failed crates after we fixed the constant (in case four)
-could compare the full second calibration in case three
-But we have to delete the temporal db after we loaded the constants!